Dec. 2009 / Vol. 128 MITSUBISHI ELECTRICscreen. Energy is also saved because the function controls...
Transcript of Dec. 2009 / Vol. 128 MITSUBISHI ELECTRICscreen. Energy is also saved because the function controls...
• Editorial-Chief
Kiyoshi Takakuwa • Editorial Advisors
Chisato Kobayashi Kanae Ishida Makoto Egashira Koji Yasui Hiroaki Kawachi Masayuki Masuda Akio Toda Tetsuyuki Yanase Tetsuji Ishikawa Taizo Kittaka Keiji Hatanaka Itsuo Seki Kazufumi Tanegashima Kazumasa Mitsunaga
• Vol. 128 Feature Articles Editor
Takenori Sakashita • Editorial Inquiries
Makoto Egashira Corporate Total Productivity Management & Environmental Programs Fax +81-3-3218-2465
• Product Inquiries Keiji Hatanaka Living Environment & Digital Media Equipment Group Fax +81-3-3218-2986
Mitsubishi Electric Advance is published on line quarterly (in March, June, September, and December) by Mitsubishi Electric Corporation. Copyright © 2009 by Mitsubishi Electric Corporation; all rights reserved. Printed in Japan.
CONTENTS
Technical Reports
Overview ..........................................................................................1 by Kenjiro Kime Image Quality Improvement Technologies for Energy-Saving LCD TVs ...........................................................................................2 by Yoshitomo Nakamura and Kohei Nomoto New Optical Disc Technologies for Blu-ray Discs .......................6 by Nobuo Takeshita, Kazuhiko Nakane and Tomo Kishigami UI Software Development for Digital Recorders with the UI Design Tool “NINA” ...................................................................................10 by Nobutoshi Todoroki and Yoshihiro Yamaguchi Laser TV .........................................................................................14 by Kuniko Kojima and Akihisa Miyata
MITSUBISHI ELECTRIC ADVANCE
Dec. 2009 / Vol. 128
Display and Storage Technology
*Living Environment & Digital Media Equipment Group Mitsubishi Electric ADVANCE December 2009 1
TECHNICAL REPORTS
Overview
Author: Kenjiro Kime*
Driven by digitalization of broadcasting, increasingly high-definition packaged media, and changes and diversification in video-display devices, all types of video/information equipment from home use to business use are undergoing dramatic changes. As a result, there is intense competition to improve the digital image quality and performance of display devices. Also, advances in networks and integration of broadcasting into communication technology are under way. Our imaging technologies, which are based on our laboratory’s basic research integrated into commercial home-use equipment, have been developed for many years and are also found in business-use equipment. To improve digitalization and image quality, we have focused not only on simple image processing but also diverse fields such as differentiating our products in each display device category diversified and our storage devices. In Japan, the shift to digital terrestrial broadcasting will be completed in about 18 months. Although the process has been relatively smoothly so far, it is important to complete it by the target date. Therefore, a plan of comprehensive measures, led by the Ministry of Internal Affairs and Communications, was launched last year. Efforts such as setting up support centers, plans for collective housing, satellite safety net, and support for families on public assistance are being made nation-wide. Since the shift to digital terrestrial broadcasting will have a significant impact on various video-related equipment, successful completion is important.
*Advanced Technology R&D Center **Yamagata University 2
TECHNICAL REPORTS
Image Quality Improvement Technologies for Energy-Saving LCD TVs
Authors: Yoshitomo Nakamura* and Kohei Nomoto**
Introduction
We have developed an energy-saving, easy-to-view brightness control technology: the lumi-nance of the liquid crystal display (LCD) TV screen is controlled to match the visual characteristics of the viewer. This technology has been adopted for the REAL MZW and REAL MXW series LCD TVs in the applica-tion of the Katei Gashitsu mode and energy-saving Katei Gashitsu mode.
1. Brightness Control under Household
Conditions In the 2006 REAL MX60 series LCD TVs, Mitsubi-
shi Electric has introduced a brightness control function, “Katei Gashitsu mode”, which eliminates dazzle. This function is the result of research on actual viewing environments, studying the visual characteristics of viewers, and surveying the performance of display devices. The function automatically adjusts the lumi-nance and hue based on the viewer’s age, viewing distance and characteristics of images displayed on the screen. Energy is also saved because the function controls brightness and reduces wasted energy.
1.1 Viewing environment
In 2004, we conducted a survey on lighting condi-tions in general households and found the average screen illuminance was 108 lx, while Nakata et al. in 1994 reported(1) a value of 93 lx, thus the screen illu-minance in an average lighting environment is ap-proximately 100 lx. The Katei Gashitsu mode was set based on a subjective evaluation under these lighting
conditions.
1.2 Black image on LCD panel Even if an LCD TV displays a black screen, the
backlight is glowing, so a very black scene may appear gray due to light leaking from the backlight. Therefore, we evaluated the relationship between the screen illuminance and the required black level (Fig. 1), and using the results, control the brightness to match the black level in the Katei Gashitsu mode.
In a bright environment, surface-reflected light will degrade the reproducibility of black. Therefore, since the MZW series LCD TVs released in October 2007, we have applied an anti-glare coating to LCD panels, which suppresses surface reflection and internal diffuse light to enhance the reproducibility of black.
1.3 Dazzling luminance
On LCD TVs without brightness control, the maxi-mum brightness is constant regardless of the displayed image. Therefore, an image with higher average brightness level will seem more dazzling than an image with a lower level.
We conducted an experiment to evaluate the daz-zling luminance for both elderly (60 to 70s) and young persons (20s). The evaluation method used circular patterns of different display sizes to determine the luminance at which the subjects started to sense daz-zling. The evaluation showed that dazzling was sensed at lower brightness when the area of the displayed pattern was larger, and under dimmer lighting dazzling was also sensed at lower brightness (Fig. 2).
Dark room environment
Home environment
Shop environment
1000 100 10 1 0.1
1
10
Dem
ande
d ba
ck le
vel (
cd/m
2 )
Screen illuminance (lx)
Fig. 1 Relation between screen illuminance and preferred black level
1 10 100
10000
1000
100Daz
zlin
g lu
min
ance
(cd/
m2 )
Adaptaion luminance (on wall) (cd/m2)
Fig. 2 Adaptation luminance and dazzling luminance
Mitsubishi Electric ADVANCE December 2009 3
TECHNICAL REPORTS
1.4 Viewing angle and brightness Figure 3 shows the relationship between the view-
ing distances at screen illuminances of 30, 100 and 300 lx, and the preferred luminance. The shorter the viewing distance, the larger the viewing angle and the lower the preferred luminance. The survey results(2) of viewing distances of adults and children revealed that children in their teens watched TV at a distance of 0.8 times that of adults. Therefore, since the MX60 series LCD TVs, the luminance is set for the viewing distance according to the setting of the illuminance sensor provided with 3-step intensity differences.
Pref
erre
d lu
min
ance
:
Pea
k w
hite
lum
inan
ce (c
d/m
2 )
Fig. 3 Viewing distance and preferred luminance
1.5 Color temperature and feeling of brightness
The setting of color temperature of household TVs varies greatly, ranging from 7,000 to 20,000 K. The pre-ferred color temperature of TV images depends on the age of the viewer and the preferred luminance. Figure 4 shows the relationship between the preferred luminance and color temperature of 20 elderly subjects. The figure shows that the higher the value, the stronger the prefer-ence. Elderly people generally tend to like images with higher color temperature because the spectral transmit-tance of crystalline lenses changes with age, but these results showed that subjects who liked higher luminance tended to like higher color temperature, and vice versa. The results are reflected in the Katei Gashitsu mode of the 2008 MXW200 series LCD TVs.
2. Brightness Control for Energy Saving and Image Quality
Figure 5 shows the relationship between bright-ness feeling and display luminance. The Katei Gashitsu mode controls luminance in the “bright” area where dazzling is not felt. The energy-saving Katei Gashitsu mode installed in the 2008 MZW200 series LCD TVs evaluates the brightness when viewers do not feel that a household environment is dark, and controls lumi-nance in this area.
Fig. 5 Peak white luminance and brightness feeling
2.1 Permissible lower bound and preferred lumi-
nance evaluation To evaluate the appropriate luminance area, a ref-
erence value is determined using the following two evaluation indices: (1) Preferred luminance: Brightness at which viewers
do not feel eye fatigue, even when viewing for a long time, and can continue to watch the high-quality images
(2) Permissible lower luminance: Brightness that is permissible when energy saving is taken into con-sideration The evaluation was performed using a 46-inch type
LCD TV in a household environment, and with 20 sub-jects aged 20 or older. Figure 6 shows the results: The preferred luminance was between 50 cd/m2 and 500 cd/m2, and the permissible lower luminance was be-tween 20 cd/m2 and 250 cd/m2.
Sub
ject
ive
eval
uatio
n of
pr
efer
red
colo
r tem
pera
ture
Preferred luminance (cd/m2) (senior (60 ∼ ): 20 people
-1.5
-1
-0.5
0
0.5
1
1.5
0 200 400 600 800
5000K6500K9300KA low color temperature
is better. A high color temperatureis better.
-1.5
-1
-0.5
0
0.5
1
1.5
0 200 400 600 800
5000K6500K9300KA low color temperature
is better. A high color temperatureis better.
Fig. 4 Relation between preferred luminance and sub-
jective evaluation of preferred color temperature
0
5
10
15
20
10 100 1000Peak white luminance (cd/m2)
Freq
uenc
y (%
)
Preferred luminancePermissible lower luminance
0
5
10
15
20
10 100 1000Peak white luminance (cd/m2)
Freq
uenc
y (%
)
Preferred luminancePermissible lower luminance
Fig. 6 Distribution of preferred luminance and
permissible lower luminance
4
TECHNICAL REPORTS
Considering these results, we chose a luminance that 90% or more of the subjects did not think was dark, and a display luminance that 50% or more of subjects pre-ferred, for the setting values. With these settings, the annual electricity consumption of the MZW200 series LCD TVs was 168 kWh/year for the 40-inch type and 185 kWh/year for the 46-inch type, among the most energy-efficient in the industry.
2.2 Settings for image quality and energy saving
Viewers generally prefer higher luminance, so the key when choosing an appropriate luminance setting is to maintain high quality image at an appropriate display luminance. This luminance level for high quality images was obtained from experimental data and theoretical analysis.
2.3 Ordered Stimuli Multi-dimensional Image Qual-
ity Optimization (OSMIQO) People’s preference for image quality is based on
fixed regularity, which is common to many of them(3). Therefore, to achieve an image with an appropriate luminance area, the regularity of image quality was obtained by a subjective evaluation. People other than TV engineers were selected as subjects, and the Or-dered Stimuli Multi-dimensional Image Quality Optimi-zation (OSMIQO) method was developed and used, in which subjects watched multiple images of different quality and evaluated them. The following two ap-proaches were selected in this method in order to compare many combinations of image quality parame-ters within a small number of evaluations: (1) Two or more images were compared in an evalua-
tion. (2) To simplify the evaluation, only one parameter of
image quality varied among the shown images; the other image quality parameters remained fixed. The images to be compared were lined up in order of the varying parameter’s value and presented to the subjects.
2.4 Experimental method For this experiment, black level, gamma and color
temperature were selected as the image quality parame-ters to be adjusted. Figure 7 shows the correspondence between the experimental scene and image quality pa-rameter space. The images used for the experiment in-cluded sports, news, a variety show, and a movie.
The subjects selected their most preferred image quality of those presented, and then were asked to rank the remaining inferior image qualities according to five grades, comparing them with the most preferred one.
2.5 Theory of analysis
We should compute the entire potential distribution
of preference in the image quality parameter space, based on the experimental data that covers the relation between some of the image qualities.
A new algorithm was needed to compute this po-tential distribution, since experimental data was limited. To construct this algorithm, the following three points must be considered: (1) The optimum image quality may not be just
one—there may be two or more optimal points within the space.
(2) Generally, there are multiple paths from an arbi-trary point to the optimal point within the space.
(3) A contradiction may occur in human evaluation, and there is a possibility of returning to the original image quality when better quality is tracked. Consequently, the OSMIQO method was devel-
oped as this algorithm.
2.6 Analysis results Figure 8 shows the potential distribution of prefer-
ence during a variety show program—the three subfig-ures show the evaluation function of the preference vs. color temperature and gamma when the black level is fixed. As can be seen from these subfigures, the lower black level setting was preferred. The same result holds true for other contents, which revealed that people truly demand deep black, even during a low-luminance setting.
In addition, the gamma and color temperature af-fect each other, showing a strongly non-linear potential distribution when the black level is low. These results are the tacit rules for the preference of image quality.
Based on these results, the image quality of the MZW200 series LCD TVs was specified. Thus, the series achieves industry-leading energy saving while offering an image quality that a majority of people prefer.
3. Conclusion
The luminance level must be set to an appropriate brightness to reduce the power consumption of LCD TVs. However, people must still like the image quality, so we will use the energy-saving Katei Gashitsu and OSMIQO developed in this study to improve image
Fig. 7 Experimental scenes and image quality
parameter spaces
Mitsubishi Electric ADVANCE December 2009 5
TECHNICAL REPORTS
quality and save energy, all at the same time, in various ways in the future.
References (1) Y. Nakata and K. Yamamoto, “The Viewing of
Television at Home,” Technical Report of Institute of Television Engineers of Japan, pp. 18, 21, 7-12 (1994)
(2) S. Kubota, M. Yamakawa, Y. Nakamura, K. No-moto and E. Kido, “Television Viewing Distances of Children,” Journal of the Institute of Image Informa-tion and Television Engineers, p. 61, No. 1 (2007) 3
(3) K. Nomoto, E. Kido, Y. Nakamura and Y. Haya-kawa, “Kansei Model of the Relation between TV Program and Desired Picture,” Journal of Human Interface Society, Vol. 9, No. 1, p. 57-69 (2007)
-1.2-0.8-0.4
00.40.81.21.6
Color temperature Gamma
Black Level=30
lowmiddle high
middle
high
off
Eva
luat
ion
func
tion
-1.2-0.8-0.4
00.40.81.21.6
Color temperature Gamma
Black Level=30
lowmiddle high
middle
high
off
Eva
luat
ion
func
tion
lowhighmiddle
middlehigh
off
-1.2-0.8-0.4
00.40.81.21.6
Eva
luat
ion
func
tion
Color temperature
Black Level=0
Gamma
lowhighmiddle
middlehigh
off
-1.2-0.8-0.4
00.40.81.21.6
Eva
luat
ion
func
tion
Color temperature
Black Level=0
Gamma
lowhighmiddle
middlehigh
off
-1.2-0.8-0.4
00.40.81.21.6
Color temperature
Black Level=-10
Eva
luat
ion
func
tion
Gamma
lowhighmiddle
middlehigh
off
-1.2-0.8-0.4
00.40.81.21.6
Color temperature
Black Level=-10
Eva
luat
ion
func
tion
Gamma
Fig. 8 Correspondence of evaluation image and image quality parameter space (case of variety show program)
*Advanced Technology R&D Center 6
TECHNICAL REPORTS
New Optical Disc Technologies for Blu-ray Discs
Authors: Nobuo Takeshita*, Kazuhiko Nakane* and Tomo Kishigami*
Introduction
Mitsubishi Electric has developed a system that automatically optimizes the write strategy for optical discs in order to shorten the development period of Blu-ray Disc (BD) recorders and to ensure stable per-formance. We have also developed tools for verifying logical specifications, which verifies the disc logical information recorded by a BD recorder or created by a BD content authoring system.
1. Technology for Automatic Optimization
of Optical Disc Write Strategy 1.1 Write strategy for optical discs
A BD recorder performs laser light modulation, which is called the write strategy, thereby precisely cre-ating recording marks according to the recording data length, in order to correctly record data on a BD disc. The write strategy involves disc-specific optimal adjust-ments for optical discs. A typical optical disc recorder pre-stores many types of optimal write strategies in its memory for use when recording. Since the optimal write strategy differs depending on the optical specifications of the optical pick-up and the discs, optimal write strategies for all discs must be prepared when the optical pick-up is modified due to a change in model of the optical disc recorder.
Figure 1 illustrates a write strategy for BDs. The BD uses recording data with 2T to 9T (where T indicates the channel clock period), and creates marks according to each recording data, which requires the laser emitting pattern to be controlled for the multi-pulse type.
1.2 System for automatic optimization of write
strategy We have developed a system that automatically
optimizes the write strategy, which depends on the characteristics of an optical disc, using a DVD recorder. Figure 2 is a block diagram of the system. The write strategy is optimized by analyzing the playback signals obtained by test recording with a DVD recorder using a digital oscilloscope and computer, and by repeating the closed loops in which the correction amount in the write strategy is fed back to the DVD recorder. The optimiza-tion for analyzing playback signals and adjusting the write strategy amount was efficiently achieved by the operators’ optimization skills and expertise.
Specifically, the optimization process starts with the step in which the initial write strategy is set in a DVD recorder, and then 8/16-modulated random data are recorded on a DVD disc. Next, the recorded data is played back, and the playback signals after equalization (EQ) that are output from the DVD recorder are stored in a digital oscilloscope. After this process, the playback waveform data after A/D conversion are stored in a computer. The playback waveform data are processed at the playback signal analysis and write strategy cal-culation blocks. The playback signal analysis block extracts jitter and several waveform parameters from stored playback waveform data; then, based on these parameters, the write strategy calculation block deter-mines the next write strategy setup values. These
Fig. 1 Write strategy for BD
Fig. 2 Block diagram of automatic write strategy
optimization system
Mitsubishi Electric ADVANCE December 2009 7
TECHNICAL REPORTS
processes are repeated until the jitter and waveform parameters satisfy the target values, and finally the optimal write strategy is determined.
Figure 3 shows the processes in the playback sig-nal analysis and write strategy calculation blocks. The playback signal analysis block calculates the signal widths (mark and space widths) based on the slice level from the playback signal data. For mark and space widths, average values are calculated from each re-cording data length. The peak level corresponding to the recording data with 3T is extracted from the play-back signal. The write strategy calculation block selects the parameters of the write strategy required for correc-tion based on the waveform parameters obtained at the playback signal analysis block, and calculates the corrected values. When the jitter and waveform pa-rameters satisfy the target values, the automatic opti-mization system outputs the optimal write strategy.
1.3 Evaluation of characteristics
In order to examine the performance of the newly developed automatic write strategy optimization system, the jitter performance and optimization time were evaluated using 16 types of commercially available DVD-R discs produced by different manufacturers. Figure 4 shows the jitter performance obtained by the
automatic optimization system. As this figure indicates, all discs showed acceptable jitter values, which were lower than the target value of 8%. Figure 5 illustrates playback signals before and after the optimization: (a) shows the playback signal when recording is made using the initial write strategy; (b) shows the playback signal recorded using the optimized write strategy. Lack of sharpness is shown in (a), which indicates the varia-tion at the portions where waveforms are overlapped. Such variation is caused by the influences of in-ter-symbol interference and waveform distortion be-cause the write strategy is not optimized. On the other hand, in (b), the variation is remarkably eliminated at the portions where waveforms are overlapped. The time required to automatically optimize the write strategy in this system is shown in Fig. 6. As can be seen, the optimization takes from 45 seconds to 2 minutes, 30 seconds. In addition, an automatic write strategy opti-mization system for BDs is developed by replacing the recorder block of the system with an evaluation equip-ment that can perform recording on and playback of BD recorders to customize the optimal algorithms for BD’s. 2. Structure of BD Specifications and
Formal Verification Tools 2.1 BD specifications
Figure 7 shows the structure of BD specifications. The specifications are classified into three layers. The specifications for the physical layer consist of specifica-tions for rewritable discs, recordable discs, and read-only discs. These three specifications define the type of optical disc and the characteristics of its re-
Fig. 3 Signal analysis and write strategy calculation
Fig. 4 Jitter value with optimized write strategy
Fig. 5 Playback signal of before and after optimization write
strategy
Fig. 6 Time for write strategy optimization
8
TECHNICAL REPORTS
cording and playback signals. To maintain and improve the recording signal quality of optical discs, which are interchangeable media, the disc characteristics must conform to those of the recorder. To ensure compatibility between various types of commercially available discs and recorders, it is necessary to secure the quality levels between them. The specifications licenser, BDA, dictates verification tests to the manufacturers to assure compati-bility, and only those products that pass the test are certified as BD products. This verification test includes two kinds of tests. One is a confirmation of the signal quality of a disc when it is recorded by a reference re-cording device; the other is a confirmation of the re-cording signal quality of a recorder when it records data on a test disc having the qualified characteristics.
The file system and application layers are collec-tively referred to as logical layers, which define the alignment and meaning of all data on the optical disc, which include video content to be recorded as digital data files or file management data for them. For the recorder that records video content, the digital video data format recorded on a disc must conform to the logical specifications. In the verification test, test data is recorded on a disc and the disc is analyzed with the formal verification tool, called a verifier, to check whether the data format and alignment satisfy all defini-tions and conditions in the specifications. There are three specifications for the file system layer, corre-sponding to each physical specification. The application layer also has three specifications: the Blu-ray Disc Audio/Visual specifications (BDAV) for broadcast pro-gram recording, the Blu-ray Disc Movie specifications (BDMV) for movie content, which realizes various video presentations, and the specifications for using the
BDMV specifications for personal recording. Verifiers were prepared for each specification.
Compatibility of BD discs recorded by one BD re-corder among the other BD recorders or players is assured by confirming that all products satisfy condi-tions of the specifications in the formal verification test.
2.2 Formal verification tool
Figure 8 shows the configuration of the BDAV specifications verifier used for the recorder verification test, as an example of the formal verification tool. Digi-tized video data are classified into the stream data group that represents video and audio for a certain period of time, and the control database group that manages and controls the stream data including the connecting relations and replay conditions of each stream data. Each of them is recorded on the disc as a separate file. The specifications define the detailed format of these data, and thus, the recorder shall create data files so that all conditions are satisfied upon re-cording on a disc.
The verifier, which is a software program that runs on a personal computer, reads recorded data on a disc through a BD drive installed in the personal computer to conduct a test. The verifier is used to inspect whether or not the contents of each file above defined in the specifications conform to the specifications and to precisely inspect whether or not the contents in the management database are consistent with the contents of the stream data groups and their allocation on the disc. If nonconformance to the specifications or incon-sistency among data is found, information about the conformance error to the specifications is reported.
BD-RE (Part1) ver.2 BD-ROM (Part1) ver.1BD-R (Part1) ver.1
BD-RE (Part2) ver.2 BD-ROM (Part2) ver.1BD-R (Part2) ver.1
BD-RE (Part3)ver.2 BD-ROM (Part3) ver.2BD-RE (Part3)
ver.3
Logical SpecificationsLayer BDMV
SpecificationsBDMV
SpecificationsSame Specificationsas for the Rewritable
disc is applied.
Physical SpecificationsLayer
(Part1)
ApplicationSpecifications
(Part3)
File systemSpecifications
(Part2)
BDAVSpecifications
Recordable Disc Read-only DiscRewritable DiscDisc Type
Official verification tool (Verifier) for each specifications, approved by BDA
BD
BD-RE (Part3)ver.2 -- E
-
...
BD-RE (Part1) ver.2 BD-ROM (Part1) ver.1BD-R (Part1) ver.1
BD-RE (Part2) ver.2 BD-ROM (Part2) ver.1BD-R (Part2) ver.1
BD-RE (Part3)ver.2 BD-ROM (Part3) ver.2BD-RE (Part3)
ver.3
Logical SpecificationsLayer BDMV
SpecificationsBDMV
SpecificationsSame Specificationsas for the Rewritable
disc is applied.
Physical SpecificationsLayer
(Part1)
ApplicationSpecifications
(Part3)
File systemSpecifications
(Part2)
BDAVSpecifications
Recordable Disc Read-only DiscRewritable DiscDisc Type
Official verification tool (Verifier) for each specifications, approved by BDA
BD
BD-RE (Part3)ver.2 -- E
-
... Fig. 7 BD specifications and formal verification tools
Mitsubishi Electric ADVANCE December 2009 9
TECHNICAL REPORTS
Fig. 8 Functions and features of verifier
The BDAV specifications employ both Moving Pic-
ture Experts Group (MPEG)-2 and MPEG-4 Advanced Video Coding (AVC) High Profile. The verifier supports these video compression technologies. The stream data are analyzed in a unit of Group of Pictures (GOP), and conformance to the relevant parameter of the video compression technology standard, which are referred to in the BD specifications, is also inspected. The devel-oped verifier reduces the verification test time by using multi-thread processing for verification operations, and features viewer functions that interactively display management database syntaxes and stream data structures. These features make it easier to detect recorder design failures.
3. Conclusion
We have developed a system that automatically optimizes the write strategy for recorders and drives. We have also developed formal tools for verifying logi-cal specifications for the verification test that checks whether BD products conform to the BD specifications, which are used by BD testing centers.
*Advanced Technology R&D Center **Kyoto Works 10
TECHNICAL REPORTS
Fig. 1 The structure of NINA and the UI development flow Fig. 2 Hierarchical use of SCOs
UI Software Development for Digital Recorders with
the UI Design Tool “NINA” Authors: Nobutoshi Todoroki* and Yoshihiro Yamaguchi**
Introduction
The cost of developing user interface (UI) software has continued to rise as embedded devices become more functional and UI designs become richer. We have been developing the UI design tool “NINA (Navigator for Interface of Application)” to improve the efficiency of UI development for such embedded devices and reduce the cost. This report discusses the application of NINA to the UI development of a digital recorder and its successor models, and evaluates its effectiveness.
1. UI Design Tool “NINA”
“NINA” is a UI design tool for embedded devices; it consists of several functional modules used on PC and a software module for target device (Fig. 1). Although embedded devices have numerous functions and set-ting items, the screen size is often limited. Therefore, the UI for embedded devices is generally operated by switching the screens for each particular procedure. “NINA” is used to design the UI based on the concept of SCO (State Chart Object), which is useful for modeling UI involving such switching of screens.
1.1 SCO
SCO is a UI component that consists of more than one state and transitions between states. Each state has one scene layout. The scene layout that corre-sponds to the SCO state at a certain point is displayed as the external appearance of the SCO at that point.
In each scene layout, other SCOs can be laid out as UI components in addition to the basic UI compo-nents, such as buttons and labels. This enables the SCOs to be made hierarchical, which allows the de-signer to design all custom UI components on screen to the UI of device application in the same framework, and then operate them by combining them.
Figure 2 shows an example of the hierarchal use of SCOs. SCO1 has two states – stop and play of music – and when a button is pressed, the device transfers to the other state. In the scene layouts of each state, the UI screens that correspond to “stop” and “play” are created. SCO2 achieves the UI that displays the play-back screen when a track to be played is selected. On the playback screen, SCO1 is laid out in the scene layout as a custom UI component. Modifying the design of SCO1 in such designing also automatically changes the playback screen of SCO2.
1.2 Components
An overview of the major components of NINA is given in Fig. 1. The chart editor is a tool for designing the transitions between the UI screens as state dia-grams. It also describes the procedures in response to key events during device operation and other events generated in the device, as an event handler. The event handler is written using a Java-based simple script language (Java is a registered trademark of Sun Mi-crosystems, Inc.).
Mitsubishi Electric ADVANCE December 2009 11
TECHNICAL REPORTS
Fig. 3 A setting menu
The layout editor creates a scene layout in each state of the SCO. A scene layout combines the SCOs and the basic UI components, including labels, buttons, images, and containers that group the UI components.
The simulator is a tool which simulates the UI op-eration designed using the chart editor and layout editor, allowing the designer to check it on a PC used for development.
The code generator uses the UI design data to generate C++ source code for target devices: The generated codes do not include any plat-form-dependent process and do not need to be modi-fied by the programmer.
The SCOOT (SCO Oriented Technology) runtime is a software module written in C++. It runs the source codes generated by the code generator on a target device.
2. UI Development of Digital Recorder
We used NINA for the UI development of our DX-TL6000 digital recorder for Japan. Based on the UI, we have developed UIs for six successor models such as DX-TL308E, including models for Europe. An over-view of the digital recorder and the UI development of these models using NINA is given below.
2.1 Overview of digital recorder
The digital recorder is a business-use surveillance device which is mainly embedded in the security sur-veillance system of retail stores and banks. It digitally records surveillance images for a long time.
The latest DX-TL6000 models use a USB mouse and have a simple menu so the user can operate the recorder like a PC. The features of this UI are as follows: (1) The user can operate the UI using both the front
buttons on the front panel and the USB mouse. The front buttons include the six operation buttons: up, down, left, right, enter and back. These buttons move the focus that indicates the UI component that the user is currently manipulating or scroll it, thus selecting the item on the screen and operating “enter/back” to move between the screens or set values. The user can also use the USB mouse to move the cursor and click it just like using a PC.
(2) A digital recorder has particular operations – Focus management in the vertical and horizontal direc-tions, defined in a complicated manner, and nested focus management. The following focus manage-ment is performed on the screen shown in Fig. 3:
Pressing the enter key when a block-shaped UI component is being focused will move the focus to the UI component within the block. In the block, the focus will move by operating the left and right keys.
Pressing the enter key when a UI component in a block is being focused will move the focus to the entire block.
Clicking the mouse on each UI component will di-rectly move the focus to the component.
2.2 UI development by using NINA To achieve the UI described above, the following
UI components have been developed: (1) Extending basic UI components
The basic UI components have been extended so that they can be operated by both pressing the front buttons and clicking the mouse. (2) Designing custom UI components for digital re-
corders Complicated focus management can also be
achieved by describing it using script language. How-
12
TECHNICAL REPORTS
ever, there are many similar screens, so many clones of code are generated, which increases the quantity of coding and makes it difficult to cover all changes when the specifications are to be changed. To prevent this, we offer custom UI components that include each focus management as the UI design using the framework of SCO. This component can be laid out in the same way as a basic UI component on various screens. Even when the focus management specifications are changed, simply revising this UI component will be reflected in all screens that are using the UI component.
The UI was developed by developing this UI com-ponent. As a result, a 580-screen UI was developed in 16 man-months.
2.3 Developing UIs of successor models using
NINA The UI development of six DX-TL6000 successor
models is outlined below. The major changes in UI specifications from the DX-TL6000 are as follows: (1) There are fewer cameras to be controlled and
fewer options for various settings. (2) The video signal used in the models for Europe is
different: the screen display area (in the vertical di-rection) is 1.2 times that in models for Japan. For (1), the scene layout for fewer options was also
added to the UI component that displays the setting items, and each layout is selectively displayed corre-sponding to the model. For (2), we provided a new basic UI component for which two types of layout pa-rameter can be set and the layout can then be switched for display. Simply replacing the previous basic UI component with this new one allows the display to handle two types of layout using the one screen.
Based on these change strategies, the UI of six successor models was developed. This encouraged reuse of the UI design, and the development was com-pleted by partially revising and adding the definition of UI design, taking only 2.6 man-months.
2.4 Evaluation of application of NINA
With the DX-TL6000, the types of screen increased considerably compared with previous models, and the specifications became much more complicated be-cause the mouse and operation buttons were used concurrently. Nevertheless, the UI was achieved using only half the man-hours of the previous UI. This section describes the evaluation of this achievement, and also considers reusing the UI design in developing future successor models.
2.4.1 Reducing the number of created screens by
common use One common screen that can be used as a base is
created, and UI components are then added or deleted
as necessary. This allows the designer to express many screens, and this common use of screens sub-stantially reduces the number of created screens in the UI specifications with many similar screens. We ana-lyzed the UI specifications prior to development and tried common use of screens (Table 1). The number of actually created screens was 253, so we successfully reduced the number of screens to 44% of that defined in the UI specifications.
Table 1 UI development cost of DX-TL6000
Development man-hours
No. of speci-fied screens
No. of cre-ated screens
Develop-ment scale
16 man-month 580 253
2.4.2 Increasing development efficiency by intro-
ducing custom UI components With the screen in Fig. 3, we compared the
man-hours required to describe all focus move-ment/scroll movement using a script language, and when using a custom UI component. In the latter case, the man-hours for the focus portion were reduced by 65% and the man-hours for designing the UI were reduced by 39% for the entire screen in Fig. 3. There-fore, the man-hours required for the product as a whole will be further decreased.
2.4.3 Increasing the efficiency of development by
using a simulator If a detailed test is conducted in a prior work proc-
ess using the simulator function of NINA, faults in the post-work process can be reduced. Table 2 shows the number of faults in the UI portion in the simulator test and system test: it is clear that, among the faults that can be detected in the whole work process, 90% were detected in the prior work process.
Table 2 Simulation test and system test results
No. of test items No. of detections
Simulator test 6100 487 System test 20000 53
2.4.4 Increasing the efficiency of development of
successor models When developing six successor models, we modi-
fied the UI design of the DX-TL6000, using it as the base. Table 3 shows the number of SCOs that could be reused in the development of these models. SCOs accounting for 63% of the entire design could be reused for designing the DX-TL6000; the development man-hours for the remaining 37% were 2.6 man-months.
Mitsubishi Electric ADVANCE December 2009 13
TECHNICAL REPORTS
Table 3 Number of reused SCOs and newly developed SCOs
Total No. No. of reuse
No. of new/additional creation
No. of SCO 270 170 100
3. Conclusion
The UI design tool NINA was used to develop the UI of digital recorders. Since the framework of reusing the UI design was used, the UI development efficiency was improved when developing an initial product and successor models. We will continue to develop suc-cessor models in the future.
*Advanced Technology R&D Center **Kyoto Works 14
TECHNICAL REPORTS
Laser TV Authors: Kuniko Kojima* and Akihisa Miyata**
Introduction
Mitsubishi Electric has developed the world’s first large-screen laser television system using a three primary color laser light source. This television features super-high resolution with double the color reproduction gamut and high contrast ratio compared to conventional LCD televisions. Displaying vivid and dynamic images, this large-screen television features a compact design, and will open up new horizons.
1. Background of Laser Television Devel-
opment We have commercialized the world’s first laser
television system using a three primary color laser light source. This television features double the color repro-duction gamut and high contrast ratio compared to conventional LCD televisions. Through this laser televi-sion system, we propose a new type of large-screen television with super-high resolution and innovative compact design.
The demand for large-screen televisions has been growing as high-vision broadcasting featuring high-definition images becomes increasingly popular. Moreover, since viewers tend to prefer to watch a large-screen television, the demand for large-screen, high-quality-image televisions with high definition and wide color reproduction gamut has been growing. Meanwhile, work on the worldwide standardization of extended color space, the color space that can render a wide color gamut, has been under way(1)(2). The ex-tended color space xvYCC, which is the standard for motion pictures, was internationally standardized in January 2006. Display devices have also been sup-porting wide-gamut color reproduction(3).
2. Features of Laser Televisions
Our laser television realizes an extremely wide color reproduction gamut which is double that of tradi-tional LCD televisions by employing a three primary color laser light source. This wide color reproduction gamut produces vivid images which could not be dis-played in the past.
The laser television incorporates a newly devel-oped laser light source, super wide-angle optical engine dedicated to the laser with far smaller optical system thanks to its smaller diameter by taking advantage of the laser light with small divergence angle, thin screen, small case, and compact laser driving power supply.
Thanks to the design with these compact components, the laser can be hung against a wall. Moreover, since the laser light source with high luminous efficiency is optimally driven with the lighting control circuit, the power consumption is as low as 135 W, one-third that of a conventional LCD television, even though the screen size is large.
This television includes 3D video display function, enabling both normal broadcasting and 3D images to be shown on one laser television. By wearing dedicated glasses, multiple viewers can enjoy dynamic 3D images with high color reproduction on the wide screen at the same time.
3. Development of Laser Television
We started developing laser televisions in earnest in 2005. After prototyping 52-inch and 56-inch televi-sions, we then developed and marketed 65-inch laser televisions(4) to (6). Table 1 lists the specifications of laser televisions that have been developed.
The 52-inch prototype television developed in 2006 was used to study color reproduction gamut of the laser light source and color rendering methods as a basic verification of the optics of laser light image rendering. The 56-inch prototype television developed in 2007 was used for verification of thinner product designs. Since both prototypes were intended as fundamental proto-types, the optical engine was installed in the television case, but the laser light source, laser light source driv-ing circuit, and video signal processing circuit were stored in the table on which the television rests. The laser light source and optical engine were connected through optical fiber cables.
The laser television developed and marketed this time contains all components within the case with a depth of 269 mm. For compactness, the laser light
Table 1 Specifications of laser TV
208 LaserVue
2007 Prototype
2006 Prototype
Screen size (inches) 65 56 52 Laser power (cd/m2) 500 500 500 Lawser power (W) 19.4 19.6 18
Dimensions (HxWxD) (mm) 1011x1466x269 918x1280x263 864x1260x473
Color gamut on u’v’ (%)
(vs ITU-RBT_709) 208% 182% 179%
Mitsubishi Electric ADVANCE December 2009 15
TECHNICAL REPORTS
source, laser driving power supply, and optical engine were made smaller. Moreover, a new laser module was developed to deliver a luminance of 500 cd/ m2 (@ 12,000 K) with a 65-inch television. Also, an advanced video signal processing LSI is used in this laser televi-sion. These element technologies are explained below.
4. Element Technologies of Laser Televi-
sion As shown in Table 1, an ultra-thin laser television
with screen size of 65 inches and depth of 269 mm was developed. A small optical engine with increased light utilization efficiency was also developed, which pro-duces a luminance of 500 cd/m2 (@ 12,000 K) with a 65-inch television. Figure 2 illustrates the color repro-duction gamut of the laser television represented on a u’ v’ chromaticity diagram. This laser television has double the color reproduction gamut compared to con-ventional LCD televisions.
Figure 3 shows the new optical engine for laser televisions. This engine incorporates a super wide-angle projection optical system that consists of a projection lens and aspherical mirror, which can project as large as a 65-inch screen from the aspherical mirror at a projector distance of 100 mm. A large convex aspherical mirror, which is included in this engine, minimizes color aberration and distortion, which are
typical problems in super wide-angle projection. The diameter of the optical system has been made smaller thanks to the higher directionality of laser light than lamp light, hence the optical system is smaller and the contrast is higher. Figure 4 (a) shows the newly-developed projection lens for the laser light source and Fig. 4 (b) shows the conventional one for a lamp light source (7). Since the diameter of the optical system is smaller, the maximum effective diameter of the projection lens is reduced by about 40%. This su-per-wide angle projection system realizes an optical engine dedicated to thin laser televisions with increased image quality and compactness.
Figure 5 illustrates the laser light source unit de-veloped for our laser television. To design a thin laser television, the laser module for RGB, cooler for the laser module, laser driving power supply, and control circuit are included in the unit which is just 106 mm thick. The control circuit for the laser light source re-ceives the light-up timing signal of laser from the video signal processing circuit, and drives the laser based on this signal. Each of the laser modules includes a laser and coupling optical component, and thus the laser light from the laser is input into the optical fibers through the coupling optical system. The laser light source unit and the optical engine are connected by the optical fibers, which guide the laser light illuminated at the laser light source unit to the optical engine.
The wavelengths of the three RGB primary colors used in the development are 447 nm for blue, 532 nm
Fig. 1 Laser TV
U′
V′
0.0 0.1 0.2 0.3 0.4 0.5 0.60.0
0.1
0.2
0.3
0.4
0.5
0.6
LaserTV
LCD-TV
Fig. 2 Color reproduction gamut
Fig. 3 Compact optical engine
Fig. 4 Projection lens
16
TECHNICAL REPORTS
for green, and 640 nm for red. The peak outputs are 6.3 W for red, 5.0 W for green, and 8.1 W for blue, totaling 19.4 W. This ratio is designed to give the optimal color temperature of white. Modules including the coupling optical component connected to the optical fibers that link the laser light source unit and optical engine are also downsized.
Figure 6 shows the new laser light source for green. A 5.0 W optical output of green light is successfully obtained in this single device.
If conventional video signals are rendered on a display with wide color reproduction gamut, an unnatu-ral image with darker colors is displayed due to the large difference between the color space in video sig-nals and that of the display. Conventional video signals are designed based on the color space for cathode ray tubes(8). In a display with wide color gamut, which typi-cally has an LED or laser as its light source, the differ-ence between the color space of video signals and that of the display becomes significant. If conventional raw video signals are rendered, deep colors are vividly rendered, and light colors are also rendered darker, because the narrow color space of video signals is linearly enlarged on the large color space of the display. If a memory color such as that of flesh is rendered reddish, the flesh color will look unnatural to viewers.
We have solved this problem using Natural Color Matrix (NCM), our proprietary color management tech-nique. The NCM technique allows video signals to be resolved into a luminance component and 12 color components, and then re-composed into video signals that are suitable for the color space created by laser television, through matrix operations.
Figure 7 shows the chromaticity measurements of typical colors displayed on a laser television and a liquid crystal television. The colors were sampled from several scenes in the contents for assessment. As shown in the figure, colors that are close to the original color are reproduced with greater depth on the laser television due to its large color reproduction gamut, while the flesh color plotted in the center is reproduced with almost the same chromaticity. This shows that the non-linear color mapping of NCM allows colors that are close to the original color to be reproduced more vividly, which are unique to laser television, than on conven-tional televisions, while light colors such as flesh color are reproduced similarly to those on conventional tele-visions. This means that more visually powerful images with natural and deep colors will be displayed using the conventional video signals. If necessary, NCM can reduce the color reproduction gamut of laser television to the gamut of conventional television. In addition, the newly developed signal processing system conforms to
x.v.Color (See Note 1 ), which allows high-fidelity color reproduction in large color space.
5. Summary
We have commercialized the world’s first laser television system using three primary color laser sources. By developing a new driving power supply, signal processing technology, optical engines, screens,
Note 1 A trademark that shows compliance with xvYCC (IEC 61966-2-4), an international standard for extended color space used for moving video pictures, which was approved and issued by the International Electrotechnical Commission (IEC).
Power supply of lasers
(B) Lasers
(G)
(R)
Fig. 5 Laser light source unit
Fig. 6 New green laser
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7u'
v'
LCD-TVLaser TV
Fig. 7 Color reproduction of laser TV
Mitsubishi Electric ADVANCE December 2009 17
TECHNICAL REPORTS
and cases suitable for the new laser sources, we have paved the way for large-screen televisions with 3D video display function, while achieving a compact size and low power consumption.
References (1) IEC 61966-2-1 Amendment 1, “Multimedia systems
and equipment – colour measurement and man-agement – Part 2-1: Colour management – Default RGB colour space – sRGB” (2003)
(2) “Exchangeable image file format for digital still cameras: Exif Version 2.21 (Amendment Ver2.2),” JEITA CP-3451-1 (2003)
(3) IEC 61966-2-4 First edition, “Multimedia systems and equipment – colour measurement and man-agement – Part 2-4: Colour management – Ex-tended-gamut YCC colour space for video applica-tions - xvYCC” (2006)
(4) J. Someya, Y. Inoue, H. Yoshii, M. Kuwata, S. Kagawa, T. Sasagawa, A. Michimori, H. Kaneko,
and H. Sugiura, “Laser TV: Ultra-Wide Gamut for a New Extended Color-Space Standard, xvYCC,” SID06 Digest, 1134–1137 (2006)
(5) H. Sugiura, M. Kuwata, Y. Inoue, T. Sasagawa, A. Nagase, S. Kagawa, N. Watanabe, and J. Someya, “Laser TV – Ultra Wide Color Gamut in Conformity with xvYCC,” SID07 Digest, 12–15 (2007)
(6) H. Sugiura, T. Sasagawa, A. Michimori, E. Toide, T. Yanagisawa, S. Yamamoto, Y. Hirano, M. Usui, S. Teramatsu, and J. Someya, “65-inch, Super Slim, Laser TV with Newly Developed Laser Light Source,” SID08 Digest, 854–857 (2008)
(7) M. Kuwata, T. Sasagawa, K. Kojima, J. Aizawa, A. Miyata and S. Shikama, and H. Sugiura, “Projection Optical System for a Compact Rear Projector,” Journal of the SID, 14/2, 199–206 (2006)
(8) Recommendation ITU-R BT.709-4 parameter val-ues for the HDTV standards for production and in-ternational programme exchange (1990, 1994, 1995, 1998, 2000)